For the last couple of days I’ve been revisiting an idea for a “tone control” preamp. I had originally decided to go one way with the design, but then something happened to change my mind. My ears got involved.

The idea originally presented itself to me about ten years ago. After a little prodding from some folks on the diyaudioprojects forum, I designed a James-Baxandall tone control network for inclusion in a vacuum tube project. After looking at my personal requirements, I decided to design the controls for approximately +/- 20dB of control in both the bass and treble regions. With these control points I landed on a design that, while not achieving the exact range I wanted, had significant control range for both controls. The response plots for that tone stack look like this.

I was fairly happy with this technical design. I proceeded to develop an overall topology and did some additional design work to include the effects of circuit loading and gains. This resulted in an overall preamp design with which I was pleased. I even constructed a single channel prototype of the control circuit to perform some additional testing. Here’s the prototype.

I did a little initial testing on this prototype but then the entire project got pushed aside.

However, my inaction did not stop several other people from implementing my design in a variety of forms. Most people were happy with the performance. But there was one recurring theme upon which I continued to get questions. I was getting reports from various people that this circuit was causing weird distortion when elevating the bass and treble controls too far. What constituted “too far” varied but the reported effects were always very similar.

I quickly identified the cause. The circuit was being used in preamps driving vacuum tube amplifiers with inputs biased to a few volts. When driven with line level inputs such amplifiers work flawlessly. However, in a preamp driven by a line level signal, keeping the tone controls neutral (i.e. at approximately 50% setting) resulted in a line level output signal. But when the bass or treble controls were elevated, the level of the output signal at some frequencies could rise by as much as 20dB or a factor of 10. So a line level signal of say 1V-rms suddenly had excursions at some frequencies that were as large as 10v-rms (14.1v peak). Under the right conditions, these signals were overdriving the amplifiers.

Now in my misspent youth, I did some amateur sound recording and mixing. So I knew the proper use of gain and volume controls and how to balance signals for control variation. I knew that when tone controls get adjusted above nominal, the preamp gain control needs to be reduced by a commensurate amount to maintain the line level inputs to the power amplifier. If this causes an overall volume falloff, the power amp volume control is adjusted to compensate. This way nothing is overdriven until the power amp volume gets raised significantly.

It soon became clear that those contacting me were not following this gain monitoring practice. This is not an unknown phenomenon among audio designers. For this reason, the tone controls on most consumer equipment provide a significantly smaller adjustment range than I included. Most controls on consumer equipment provide a maximum boost or cut of somewhere between 6dB (2v/v) and 12dB (4v/v). As I said above, the ranges I had included in my design were providing almost 20dB or 10v/v with frequency.

In light of all this, I had decided that if I were going to publish a design for a preamp with tone controls I should probably redesign my tone stack for some more modest control ranges. Over the last couple of days I’ve been looking at control levels between the aforementioned 6dB and 12dB. I reasoned that this should be enough for most situations and it should limit the number of people having issues with overall system performance. Then a funny thing happened.

It occurred to me that I still had the prototype I had built almost ten years ago. I thought maybe before I jump into a complete preamp redesign effort, I should revisit the old design and see what I thought. So I dug the old prototype out of my parts closet and wired up a test rig consisting of a source, the tone stack, and a 4S preamp for gain recovery. Here’s the test setup sitting on my desk.

After I cobbled together this setup, I proceeded to try it out with a couple of different amplifiers. I played with different genres of music and different listening levels and really used my ears to evaluate what I was doing. While I was listening, adjusting tone controls and fine tuning gain and volume settings, I realized two things.

First, with most music, the tone adjustments never really varied outside of approximately the 10 o-clock and 2 o-clock positions (maybe 9 o-clock and 3 o-clock). This was enough variation to adjust most music to the room and listening conditions. Second, I found that for some music (and some amps) I needed significantly more variation in bass and treble controls to achieve the tonal shaping I really wanted. Occasionally almost to the limits of the controls. The changes for most music were subtile but noticeable, but for a few tracks I really wanted that addition range of control. After realizing this I decided that a smaller control range was a mistake.

I’ve decided to stick with the original tone stack design. I’ll build it as a full preamp with multiple inputs and dual gain controls. This should give me all the flexibility I need. When I finally post the project page I will add a tutorial on gain levels and adjustments to maintain overall performance. I just can’t see restricting the tone control range when I really want it.

This project is still a ways off. I have the Marblewood 2.0, the 807 triode amp, and a KT88 triode amp all in the queue already. But you never know how priorities could shift in the coming year.

As always, questions and comments are welcome.

By the way, Happy Spring! At least to those of my readers in the northern hemisphere. Happy Fall to those in the southern hemisphere.